topic How to Deal with Imbalanced Multiclass Datasets with Model Studio? in SAS Data Science

How to Deal with Imbalanced Multiclass Datasets with Model Studio?

alisio_meneses — Tue, 28 Feb 2023 11:05:55 GMT

Hello there.

I am using SAS Viya 3.5 and, as far as I know, Model Studio can deal with imbalanced binary data. It is possible to set the desired proportion of event/non-event using the Event-Based sampling option in the Project Settings. But what about imbalanced multiclass data? Say I have a target with seven classes, some rare, some not. How to deal with this scenario in Model Studio?

Thanks in advance.

Re: How to Deal with Imbalanced Multiclass Datasets with Model Studio?

sbxkoenk — Mon, 06 Mar 2023 16:04:07 GMT

Paper 4051-2019
Asking the Right Questions. A Study in Multiclass Imbalanced Classification for H1-B Visas
Matthew Bunker and Owen Kelly, Kennesaw State University
https://www.sas.com/content/dam/SAS/support/en/sas-global-forum-proceedings/2019/4051-2019.pdf

This was done using Enterprise Miner.
Enterprise Miner is the predecessor of Model Studio for VDMML
(VDMML = Visual Data Mining and Machine Learning)

I haven't read it though 😞

Koen